Hi Linus:
This push fixes a bug in mv_cesa that causes all hash operations
that supply data on a final operation to fail.
Please pull from
git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6.git
or
master.kernel.org:/pub/scm/linux/kernel/git/herbert/crypto-2.6.git
Phil Sutter (1
Hi Linus:
Dave reported that even with the latest stack bloat reduction
fix for i386, the stack usage of sha512 still exceeded 1024 bytes
on sparc32. However, with Alexey's ror64 patch applied, the stack
usage goes back down to a more respectable 824.
So I've moved the ror64 patch from cryptodev
Hi Linus:
This push updates the sha512 fix so that it doesn't cause excessive
stack usage on i386. This is done by reverting to the original
code, and avoiding the W duplication by moving its initialisation
into the loop.
As the underlying code is in fact the one that we have used for years,
I'm
On Wed, Jan 25, 2012 at 8:07 PM, Herbert Xu wrote:
>
> Oops, I had incorrectly applied the first patch in the thread.
>
> I've fixed it in the tree now.
Oh well, I already pulled your tree. I just wanted to voice a few
comments on it.
>> We also avoid the problem with absolutely horrible gcc reg
On Wed, Jan 25, 2012 at 07:35:19PM -0800, Linus Torvalds wrote:
>
> Ugh. This once more has the crazy signed integer modulus operator,
> which can be quite expensive depending on whether the compiler can
> tell whether it is always positive or not.
Oops, I had incorrectly applied the first patch i
On Wed, Jan 25, 2012 at 6:43 PM, Herbert Xu wrote:
>
> This push fixes a race condition in sha512 that affects users
> who use it in process context and softirq context concurrently,
> in particular, this affects IPsec. The result of the race is
> the production of incorrect hashes, which for IPs
